NYC Subway Accessibility-Underserviced Areas
darya akimova
October 14, 2017
Main question to address: What areas of NYC are under-serviced by ADA accessible stations?
Plan: Somehow randomly “sample” locations in the NYC territory in longitude, latitude units and then calculate distance from that point to the nearest accessible subway station. Normalize to nearest (accessible or not accessible) station?
Need: - data for which stations are ada accessible - data for station locations (longitude, latitude) - some way to sample the territory of NYC (not including Staten Island and all the rivers)
library(tidyverse)
library(ggmap)
library(stringr)
library(cowplot)
library(viridis)
library(geosphere)Not sure which are the most reliable datasets to use here.
ada.raw.data <- read_csv(file = "../data/Elevator Escalator Station Data.csv") %>%
arrange(station)
sub.location.raw.data <- read_csv(file = "../data/NYC_Transit_Subway_Entrance_And_Exit_Data.csv") %>%
arrange(`Station Name`)
colnames(sub.location.raw.data)[3] <- "station"
colnames(sub.location.raw.data) <- colnames(sub.location.raw.data) %>%
str_to_lower() %>%
gsub(pattern = " ", replacement = ".")
colnames(ada.raw.data) <- colnames(ada.raw.data) %>%
str_to_lower()Mucking around
# What are the cols and how many missing values per col?
sub.location.raw.data %>%
is.na %>%
as.data.frame() %>%
sapply(sum)## division line station
## 0 0 0
## station.latitude station.longitude route1
## 0 0 0
## route2 route3 route4
## 848 1374 1547
## route5 route6 route7
## 1630 1741 1788
## route8 route9 route10
## 1820 1840 1845
## route11 entrance.type entry
## 1845 0 0
## exit.only vending staffing
## 1812 0 0
## staff.hours ada ada.notes
## 1828 0 1793
## free.crossover north.south.street east.west.street
## 0 29 35
## corner entrance.latitude entrance.longitude
## 32 0 0
## station.location entrance.location
## 0 0ada.raw.data %>%
is.na %>%
as.data.frame() %>%
sapply(sum)## station borough trainno equipmentno equipmenttype
## 0 0 0 0 0
## serving ada
## 0 0full.join.mess <- ada.raw.data %>%
full_join(sub.location.raw.data, by = "station") %>%
arrange(station)
inner.join.mess <- ada.raw.data %>%
inner_join(sub.location.raw.data, by = "station") %>%
arrange(station)
full.join.mess %>%
filter(!(station %in% unique(inner.join.mess$station)))## # A tibble: 2,046 x 38
## station borough trainno equipmentno equipmenttype serving ada.x
## <chr> <chr> <chr> <chr> <chr> <chr> <chr>
## 1 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 2 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 3 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 4 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 5 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 6 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 7 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 8 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 9 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## 10 103rd St <NA> <NA> <NA> <NA> <NA> <NA>
## # ... with 2,036 more rows, and 31 more variables: division <chr>,
## # line <chr>, station.latitude <dbl>, station.longitude <dbl>,
## # route1 <chr>, route2 <chr>, route3 <chr>, route4 <chr>, route5 <chr>,
## # route6 <chr>, route7 <chr>, route8 <int>, route9 <int>, route10 <int>,
## # route11 <int>, entrance.type <chr>, entry <chr>, exit.only <chr>,
## # vending <chr>, staffing <chr>, staff.hours <chr>, ada.y <lgl>,
## # ada.notes <chr>, free.crossover <lgl>, north.south.street <chr>,
## # east.west.street <chr>, corner <chr>, entrance.latitude <dbl>,
## # entrance.longitude <dbl>, station.location <chr>,
## # entrance.location <chr># need to work on thisThe Subway Entrance and Exit data seems to be good enough for an initial analysis. Hopefully the ADA column in that dataset is reliable. Need to compare this with the other datasets collected at some point, but joining is a problem. To do: - check against Wire Monkey’s joined dataset
For now, stick with the sub.location.raw.data for the rest of the analysis.
Map of all subway stations in this dataset:
# map of all the stations
nyc.map <- get_map(location = "New York City", maptype = "roadmap")
ggmap(nyc.map) +
geom_point(data = sub.location.raw.data, aes(x = station.longitude, y = station.latitude)) +
ylim(40.495992, 40.915568) + # NYC city limits latitude coordinates
xlim(-74.257159, -73.699215) + # NYC city limits longitude coordinates
xlab("longitude") +
ylab("latitude") +
ggtitle("All NYC Subway Stations\nin the Subway Entrance/Exit Locations Data")## Warning: Removed 1 rows containing missing values (geom_rect).
# Note: no stations on Staten Island - will exclude it in the following analysisNotes: - Most stations have more than one row (multiple entrances/exits) - need to compress info - Station names in the station column are NOT unique (multiple 103rd St, etc) in the entrance/exit data, but station location may be a unique consistent handle to grab all the entrances/exits associated with a station - For now, will call any station that has at least one ADA-accessible entrance/exit “accessible”, but we know that some stations are only partially accessible. Different train lines are serviced by different platforms within the same stations, and not all are fully accessible. This is where the elevator/escalator data joining may be useful for future analysis.
head(sub.location.raw.data$ada)## [1] FALSE FALSE FALSE FALSE FALSE FALSEsub.loc.ada.sum <- sub.location.raw.data %>%
group_by(station.location) %>% # seems to be the best unique key for stations in the dataset
mutate(ada.sum = sum(ada)) %>% # which stations are ada accessible?
select(division:entrance.type, ada, station.location, ada.sum) %>%
unique()
sub.loc.ada.sum %>%
ungroup() %>%
arrange(desc(ada.sum)) %>%
select(division:station, entrance.type:ada, ada.sum)## # A tibble: 585 x 6
## division line station entrance.type
## <chr> <chr> <chr> <chr>
## 1 IND 6 Avenue 47-50th Sts Rockefeller Center Easement
## 2 IND 6 Avenue 47-50th Sts Rockefeller Center Elevator
## 3 IND 6 Avenue 47-50th Sts Rockefeller Center Stair
## 4 IND 6 Avenue 47-50th Sts Rockefeller Center Stair
## 5 IND 8 Avenue 34th St Elevator
## 6 IND 8 Avenue 34th St Stair
## 7 IND Queens Boulevard Jamaica-179th St Elevator
## 8 IND Queens Boulevard Jamaica-179th St Stair
## 9 IND 8 Avenue 59th St Easement
## 10 IND 8 Avenue 59th St Elevator
## # ... with 575 more rows, and 2 more variables: ada <lgl>, ada.sum <int>sub.loc.ada.sum %>%
ungroup() %>%
arrange(desc(station)) %>%
select(division:station, entrance.type:ada, ada.sum)## # A tibble: 585 x 6
## division line station entrance.type ada ada.sum
## <chr> <chr> <chr> <chr> <lgl> <int>
## 1 IRT Pelham Zerega Av Stair FALSE 0
## 2 IND 6 Avenue York St Stair FALSE 0
## 3 IND Concourse Yankee Stadium-161st St Elevator TRUE 8
## 4 IND Concourse Yankee Stadium-161st St Stair TRUE 8
## 5 IND Concourse Yankee Stadium-161st St Door TRUE 8
## 6 IRT Jerome Yankee Stadium-161st St Elevator TRUE 8
## 7 IRT Jerome Yankee Stadium-161st St Stair TRUE 8
## 8 IND 8 Avenue World Trade Center Stair TRUE 8
## 9 IRT Flushing Woodside Av-61st St Escalator TRUE 3
## 10 IRT Flushing Woodside Av-61st St Stair TRUE 3
## # ... with 575 more rowsLooking over the result, stations with only “Stair” entrances are not ada-accessible (ada.sum == 0), but stations with “Elevator” and “Escalator” entrances have more than one ada-accessible entrance (ada.sum > 0). Will call stations with ada.sum > 0 ADA-Accessible. The total value of the ada.sum probably doesn’t matter much on it’s own. Larger stations with more entrances will have a higher ada.sum score just because of the number of entrances.
sub.loc.ada.unique <- sub.loc.ada.sum %>%
select(division:route11, station.location:ada.sum) %>%
unique() %>%
ungroup() %>%
arrange(ada.sum, station, line)
# Probably don't need to keep dragging the route information around, but who knows
sub.loc.ada.annot <- sub.loc.ada.unique %>%
filter(ada.sum == 0) %>%
mutate(ada.class = "not.ada") %>%
bind_rows(
sub.loc.ada.unique %>%
filter(ada.sum > 0) %>%
mutate(ada.class = "ada.complnt")
) %>%
arrange(station)
sub.loc.ada.annot %>%
ggplot(aes(x = station.longitude, y = station.latitude, color = ada.class)) +
geom_point(size = 3, alpha = 0.5) +
coord_quickmap()
ggmap(nyc.map) +
geom_point(
data = sub.loc.ada.annot,
aes(x = station.longitude, y = station.latitude, color = ada.class),
size = 3) +
ylim(40.55, 40.915568) + # NYC city limits latitude coordinates
xlim(-74.1, -73.699215) + # NYC city limits longitude coordinates
xlab("longitude") +
ylab("latitude") +
ggtitle("All NYC Subway Stations\nin the Subway Entrance/Exit Locations Data\nColor by Accessibility")## Warning: Removed 1 rows containing missing values (geom_rect).
Now the next questions is how do I “sample” points from the area of NYC?
The NYC map shapefile might be useful?
Borough boundary shapefile from: http://www1.nyc.gov/site/planning/data-maps/open-data/districts-download-metadata.page
NYC city limits? (same source as above)
West Longitude: -74.257159 East Longitude: -73.699215 North Latitude: 40.915568 South Latitude: 40.495992
library(rgdal)
nyc.shapefile <- readOGR(dsn = "../data/nybb_17c/nybb.shp")## OGR data source with driver: ESRI Shapefile
## Source: "../data/nybb_17c/nybb.shp", layer: "nybb"
## with 5 features
## It has 4 fieldsnyc.map.df <- fortify(nyc.shapefile)
nyc.shapefile.map <- ggplot() +
geom_path(
data = nyc.map.df, aes(x = long, y = lat, group = group), size = 0.2
) Grabbed the locations from the 2015 NYC 311 calls data - use as random sample?
nyc.311.loc <- read_csv(file = "../data/nyc_311_2015_locations.csv") %>%
filter(Borough == "BROOKLYN" | Borough == "BRONX" | Borough == "MANHATTAN" | Borough == "QUEENS") %>%
unique() %>% # a lot of calls come from an identical lat/long
filter(Longitude > -78.0) # there's one odd point out
# too many points to just plot, but here's a density plot:
nyc.311.loc %>%
ggplot(aes(x = Longitude, y = Latitude)) +
geom_bin2d(bins = 300) +
scale_fill_viridis(option = "B") +
coord_quickmap()
# can do a geom_point map with all points, but it takes a long time
# uncomment if you want to see:
#nyc.311.loc %>%
# ggplot(aes(x = Longitude, y = Latitude)) +
# geom_point(alpha = 0.2, size = 0.25) +
# coord_quickmap()Not bad, can work.
test6 <- sub.loc.ada.annot %>%
mutate(station.id = paste(station, division, line, sep = " / "))
loc.test <- nyc.311.loc %>%
slice(1:10)
test6 %>%
group_by(station.id) %>%
count() %>%
filter(n > 1)## # A tibble: 5 x 2
## # Groups: station.id [5]
## station.id n
## <chr> <int>
## 1 47-50th Sts Rockefeller Center / IND / 6 Avenue 2
## 2 Brooklyn Bridge-City Hall / IRT / Lexington 8
## 3 Canarsie - Rockaway Parkway / BMT / Canarsie 3
## 4 Chambers St / BMT / Nassau 2
## 5 Forest Hills-71st Av / IND / Queens Boulevard 2test7 <- test6 %>%
group_by(station.id) %>%
mutate(avg.lat = mean(station.latitude)) %>%
mutate(avg.long = mean(station.longitude))
test8 <- test7 %>%
select(station.id, division:station, avg.lat:avg.long, ada.class) %>%
unique()
nrow(test8)## [1] 465length(test8$station.id)## [1] 465t1 <- rbind(test8$station.id, test8$station.id)
t2 <- as.data.frame(rbind(t1, t1, t1, t1, t1), stringsAsFactors = FALSE)
colnames(t2) <- paste("stat", 1:ncol(t2), sep = "")
t3 <- t2 %>% select(1:5)
test9 <- bind_cols(loc.test, t3)
test10 <- test9 %>%
gather(key = "station.num", value = "station.id", 4:8) %>%
arrange(station.num)
test11 <- test10 %>%
inner_join(test8, by = "station.id")
test12 <- test11 %>%
mutate(dist.to.stat = distGeo(p1 = as.matrix(test11 %>% select(Longitude, Latitude)), p2 = as.matrix(test11 %>% select(avg.long, avg.lat))) * 0.000621371) %>%
mutate(loc.id = paste(Borough, Latitude, Longitude, sep = " / "))
test12 %>%
group_by(loc.id) %>%
summarize(nearest.stat = min(dist.to.stat)) %>%
ungroup() %>%
inner_join((test12 %>% select(Longitude, Latitude, loc.id)), by = "loc.id") %>%
unique() %>%
ggplot(aes(x = Longitude, y = Latitude, color = nearest.stat)) +
geom_point(size = 3) +
scale_color_viridis(option = "D")
To tackle - Separate distance calculations by borough for both the random point and the station, in case “closest” stations are across a river. Probably can keep Queens and Brooklyn together? -